From a certain point-of-view it is possible that ICE cars could get generally harder to use in a geographic area if the number of EV cars reach X%, in the term of getting fuel for the ICE car.
The "gas station" as we know it is newer than ICE cars, so people had to get their gas/petrol from somewhere. "The first purpose-built, drive-up gas station opened in 1913", and there were fuel pumps before this point. I have read that in America gas was purchased at pharmacies in the 1800s (possibly well into 1900s in more remote areas?). This was similar in New Zealand, where benzine ("petrol") was sold in 4 gallon tins from "blacksmiths, grocers, country stores, and stock and station agents".
If the demand drops enough, then a return to the "old days" might start to happen in some areas (though some places in the world have always been like this even today).
If the gas station industry in USA is anything like petrol stations are in NZ, then there is an apparent thin margin of profit that the operators of the stations can operate on. Presumably the fatter margins are being made by the "the big boys" further up the vertical. Selling snacks/drinks/etc is a way of diversifying the revenue made by the station.
The following assumes that gas stations do not pivot more into EV charging, or that even if they do that it is not enough to keep at current revenue/profit levels:
If the number of ICE cars in an area drops, then some gas stations will close up shop first as the revenue dips. Then there will be larger "no gas zones" — probably further away from the well-travelled roads/highways where you might only find a gas/petrol station near a town. If it gets any worse then highways could start to see gas stations spread out further and further.
All of a sudden "range anxiety" that EV car owners could start to apply for gas station coverage in areas. Even if there is a demand, if there is not enough of demand, then the market cannot fully address that demand.